FUNCTIONAL DOSAGE
What remains from this allele?
Measure transcript usage and decay, RNA and full-length protein abundance, catalytic activity, chromatin engagement, localization, interaction partners, and rescue.
TISSUE BIOLOGY
Where does reduced functional ASH1L dosage become limiting?
ASH1L RNA is broadly detected across human tissues, but expression alone does not identify vulnerability. The mechanistic task is to measure residual function from a specific allele and test its consequences in the relevant cell type, developmental stage, and physiological state.
DOSAGE × CELL × TIME
Tissue biology begins after gene-level haploinsufficiency—not instead of it.
ClinGen establishes that loss of one functional ASH1L copy can cause disease. It does not tell us the residual function of a particular allele or which cells are most sensitive to that reduction. Those questions require linked molecular and cell-state measurements.
CELL & TISSUE SENSITIVITY
Where does that reduction matter?
Compare relevant neural, glial, immune, epithelial, muscle, bone, endocrine, and other lineages. Broad RNA detection prioritizes contexts; it does not prove clinical involvement.
DEVELOPMENT & STATE
When does the effect emerge?
Repeat measurements across differentiation, maturation, and defined physiological conditions rather than assuming one static cell assay represents a lifetime.
WHAT THE 2026 YALE STUDY ADDS
CRISPR loss-of-function perturbation across neural progenitors, immature glutamatergic neurons, mature glutamatergic neurons, and mature GABAergic neurons shows that downstream effects vary by cell state.
WHAT IT DOES NOT YET ANSWER
It is not a graded patient-allele dosage series and does not measure residual function, tissue vulnerability, or rescue for individual ASH1L variants.
HUMAN TISSUE RNA
ASH1L RNA is broadly distributed across the displayed brain and body entries.
The 51 values below are the ASH1L values displayed in the current HPA/GTEx consensus chart. HPA’s consensus method describes 55 tissue types overall, so these 51 displayed ASH1L values should not be relabeled as the total size of that full dataset. Bars compare transcript abundance—not protein activity, variant dosage, cell vulnerability, or clinical effect.
CLASSIFICATION
Low tissue specificityBroad distribution rather than one organ-enriched RNA patternPROTEIN
General nuclear expressionProtein estimates vary by assay; RNA and protein are not interchangeableSOURCE
51-entry HPA + GTEx displayNormal-tissue reference data; cross-sectional, unweighted, and not patient-specificPUBLIC EXPRESSION CONTEXT
ASH1L is broadly expressed—including brain, but not brain-only.
HPA/GTEx consensus RNA expression shows a broad systemic signal across muscle, vascular, gastrointestinal and oral, skin, bladder, endocrine, reproductive, and lymphoid tissues.
Highest signals include brain and body tissues
Brain-labeled Non-brain
- Cerebellum34.5
- Skeletal muscle33.7
- Blood vessel27.4
- Cerebral cortex26.4
- Ovary24.5
- Retina24.2
- Basal ganglia23.9
- Colon23.9
- Tongue22.7
- Urinary bladder21.3
Brain & eye
Central nervous system, choroid plexus, and retina
- Cerebral cortex26.4
- Cerebellum34.5
- Basal ganglia23.9
- Hypothalamus19.0
- Midbrain18.6
- Amygdala21.0
- Choroid plexus7.5
- Hippocampal formation21.0
- Spinal cord17.5
- Retina24.2
Endocrine
Hormone-producing organs
- Thyroid gland17.4
- Parathyroid gland17.1
- Adrenal gland16.2
- Pituitary gland17.6
Respiratory & upper digestive
Airway and proximal oral–digestive tissues
- Lung15.4
- Salivary gland17.7
- Esophagus16.5
- Tongue22.7
Gastrointestinal tract
Luminal digestive organs
- Stomach17.1
- Duodenum9.2
- Small intestine17.1
- Colon23.9
- Rectum10.8
Liver, gallbladder & pancreas
Metabolic, biliary, exocrine, and endocrine compartments
- Liver12.8
- Gallbladder8.5
- Pancreas16.5
Kidney & urinary tract
Renal and bladder tissue
- Kidney13.2
- Urinary bladder21.3
Male reproductive
Gonadal and reproductive tract tissues
- Testis11.0
- Epididymis19.6
- Seminal vesicle14.5
- Prostate17.4
Female reproductive & breast
Gonadal, reproductive, placental, and breast tissues
- Vagina18.0
- Ovary24.5
- Fallopian tube17.7
- Endometrium20.7
- Cervix19.2
- Placenta9.7
- Breast18.5
Muscle & vascular
Cardiac, skeletal, smooth-muscle, and vessel tissue
- Blood vessel27.4
- Heart muscle18.5
- Smooth muscle11.9
- Skeletal muscle33.7
Connective, adipose & skin
Barrier and soft-tissue compartments
- Adipose tissue19.3
- Skin20.9
Bone marrow & lymphoid
Hematopoietic and immune-organ tissues
- Appendix8.0
- Spleen13.9
- Lymph node10.2
- Tonsil9.1
- Bone marrow11.2
- Thymus11.3
ASH1L chart checked 22 July 2026. The current ASH1L gene page displayed 51 consensus nTPM entries; HPA’s general methods describe a 55-tissue consensus dataset overall. This page reports only the 51 displayed ASH1L values and does not infer the absent entries. Values and tissue definitions may change with future releases. A detectable transcript is not evidence that a heterozygous variant has altered that tissue.
CELL-TYPE RESOLUTION
The signal spans neural, epithelial, metabolic, muscle, reproductive, vascular, and immune cells.
Human single-cell and single-nucleus atlases refine the bulk-tissue map. HPA classifies ASH1L as low-specificity across single cells and immune cells, while highlighting adrenal-cortex cells and spermatocytes in its tissue-cell analysis and a neuron-enriched signal in deep visual proteomics.
Neural & glial
Brain-region single-cell types and separate deep-visual-proteomics categories
The brain-region atlas includes neurons, astrocytes, Bergmann glia, oligodendroglial lineages, microglia, and vascular cells. Separate DVP categories include neurons, astrocytes/neuropil, and microglia/neuropil; those modalities are not one cell-level measurement.
Airway & epithelial
Ciliated cells, alveolar type 1 and type 2 cells, keratinocytes, secretory cells, and pancreatic epithelial cells
Detection identifies cell contexts for experiments; it does not establish epithelial disease in ASH1L haploinsufficiency.
Metabolic & renal
Hepatocytes, pancreatic islets, glomerular cells, proximal and distal tubules, and collecting ducts
Human atlas detection is broad. Direct germline loss-of-function mechanisms in these lineages remain largely untested.
Muscle & vascular
Cardiomyocytes, skeletal myofibers, capillaries, smooth-muscle cells, and vascular support cells
Human expression is supported; direct functional evidence is strongest for myoblast fusion in a mouse/cell model.
Reproductive
Granulosa cells, oocytes, spermatocytes, and reproductive-tract cell populations
HPA predicts enrichment in spermatocytes and adrenal-cortex cells. A mouse fetal-ovary overexpression study is a dosage-perturbation model, not a germline loss-of-function result.
Immune
Neutrophils, macrophages, CD4 and CD8 T cells, B cells, dendritic cells, and other blood lineages
Human immune-cell RNA has low specificity. Mouse macrophage and T-cell studies supply direct lineage mechanisms, but not a human ASH1L immune phenotype.
GLIA × EPITHELIUM × STATE RESOLUTION
A response can begin normally and still fail to resolve or close.
Astrocytes and other glia help regulate neurotransmitter and ion homeostasis, metabolic support, myelination, neurovascular coupling, and inflammatory tone. ASH1L is detected in neural and glial contexts, but astroglial development and function remain substantially less resolved than neuronal effects.
Glia and barrier epithelium are priority interface models—not established “primary victims” and not an explanation assigned to an individual person.
TWO INTERFACES · ONE COMPARATIVE QUESTION
Astrocytes and mucosal epithelium perform parallel homeostatic work in different tissues.
The comparison is functional, not lineage equivalence. Both can regulate fluid and ions, metabolic support, barrier behavior, repair, and local immune tone; each must be tested in its own biological context.
Astrocytes and other glia
Prioritize transmitter and ion handling, metabolic exchange, neurovascular coupling, inflammatory shutoff, and neuron–glia recovery after a defined challenge.
Evidence boundary:Human expression and an active astrocyte research aim support testing. They are not a completed ASH1L astrocyte result.
Oral, airway, and GI epithelium
Prioritize barrier integrity, transepithelial resistance, ion and fluid transport, mucociliary or repair kinetics, and local inflammatory resolution.
Evidence boundary:Ash1l-dependent epidermal homeostasis and wound repair are direct mouse evidence; mucosal epithelium remains a human-motivated model to test.
Measure the return to baseline
Compare baseline, challenge, peak, transcriptional shutoff, barrier or metabolic recovery, and repeat challenge using the patient allele, isogenic controls, and rescue.
Discriminator:Delayed recovery must be separated from an unusually large initial response, persistent trigger, infection, medication effect, or a different underlying disorder.
OPEN
Did the response start appropriately?
Measure stimulus threshold, onset latency, magnitude, cell identity, and the pathway initially engaged.
TRANSITION
Did the system move into the next state?
Test differentiation, maturation, sleep–wake transition, network-state change, and tissue-specific adaptation.
CLOSE
Did the activating or permissive state terminate?
Measure transcriptional shutoff, electrical recovery, inflammatory restraint, and the time required for a triggered program to end.
RESOLVE
Did the system clear the burden and restore baseline?
Track recovery curves, residual state, recurrence threshold, tissue repair, and whether function stabilizes after the trigger is gone.
CELL IDENTITY
Separate neurons from glia.
Compare astrocytes, oligodendroglial lineages, microglia, and relevant CSF- or vascular-adjacent cells with neurons rather than treating “brain” as one compartment.
TIME COURSE
Sample the return—not only the peak.
Baseline, challenge, peak response, shutoff, recovery, and repeat challenge are required to distinguish excessive activation from delayed closure.
FUNCTION
Measure what glia actually regulate.
Priorities include maturation, transmitter and ion handling, metabolic support, myelination, barrier or vascular coupling, and resolution of inflammatory signaling.
ALLELE & RESCUE
Keep the patient variant in the experiment.
Use allele-aware dosage, isogenic controls, neuron–glia co-culture, matched developmental state, and rescue to test whether a prolonged state is ASH1L-dependent.
DIRECT FUNCTIONAL EVIDENCE
Where ASH1L has been experimentally perturbed—not merely detected.
These studies establish that ASH1L can have lineage- and state-specific functions. Species, perturbation direction, developmental timing, and cell identity determine what each experiment can support.
01Human cell model
Human neural cell states
hiPSC-derived neural progenitors; immature and mature glutamatergic neurons; mature GABAergic neurons
DIRECT RESULT
Pooled CRISPR knockout showed that ASH1L-associated transcriptional consequences depend on developmental stage and neuronal identity.
SCIENTIFIC BOUNDARY
Complete knockout in engineered cells does not measure residual function from an individual patient variant.
Fernandez Garcia et al., 2026 ↗02Direct in vivo
Developing brain
Neural-progenitor deletion in mice
DIRECT RESULT
Ash1l loss altered cortical development, progenitor programs, myelination, growth, and later behavioral and memory measures in the tested model.
SCIENTIFIC BOUNDARY
A neural-lineage mouse deletion does not establish cell autonomy in every brain lineage or reproduce the full human disorder.
Gao et al., 2021 ↗03Direct in vivo
Prefrontal cortex
Region-specific Ash1l knockdown in mice
DIRECT RESULT
The perturbation altered promoter chromatin and synaptic-gene transcription, shifted excitation/inhibition balance, and produced seizure phenotypes.
SCIENTIFIC BOUNDARY
A regional knockdown cannot define every cell type, variant architecture, or human seizure trajectory.
Qin et al., 2021 ↗04Direct in vivo
Thalamocortical memory circuit
Targeted CRISPR perturbation in a mouse memory task
DIRECT RESULT
ASH1L was required later in the tested cascade to maintain memories over weeks, while initial memory formation was preserved.
SCIENTIFIC BOUNDARY
This is circuit- and task-specific evidence; it does not define a person’s memory profile or prove progressive decline.
Terceros et al., 2026 ↗05Direct lineage model
Macrophage / innate immune
Macrophage and inflammatory mouse models
DIRECT RESULT
Ash1l promoted A20/Tnfaip3 expression and restrained TLR-triggered NF-κB/MAPK signaling and IL-6/TNF production in the tested models.
SCIENTIFIC BOUNDARY
This does not establish immune deficiency, autoimmunity, or hidden inflammation in people with ASH1L-related disorder.
Xia et al., 2013 ↗06Direct lineage model
CD4 T-cell differentiation
Mouse T-cell polarization with human rheumatoid-arthritis correlation
DIRECT RESULT
Ash1l and lnc-Smad3 oppositely regulated Smad3-locus accessibility and induced regulatory T-cell polarization.
SCIENTIFIC BOUNDARY
The study concerns immune-lineage regulation and autoimmunity models, not ASH1L neurodevelopmental-disorder prevalence.
Xia et al., 2017 ↗07Direct lineage model
Hematopoietic stem cells
Conditional Ash1l deletion in mouse hematopoietic stem and progenitor cells
DIRECT RESULT
Ash1l supported adult stem-cell quiescence, long-term trilineage hematopoiesis, and Hox-gene expression in the tested system.
SCIENTIFIC BOUNDARY
Stem-cell lineage biology does not establish blood-count abnormalities, immune deficiency, or leukemia risk in ASH1L-related disorder.
Jones et al., 2015 ↗08Direct tissue model
Epidermis / keratinocytes
Hypomorphic Ash1l mouse skin
DIRECT RESULT
Ash1l disruption altered keratinocyte proliferation–differentiation balance, epidermal stratification, and wound re-epithelialization.
SCIENTIFIC BOUNDARY
This supports a skin-homeostasis mechanism in mice; it does not establish a universal human skin phenotype.
Li et al., 2017 ↗09Cross-tissue model
Neuronal–cutaneous inflammatory axis
Ash1l perturbation in a mouse psoriasis model
DIRECT RESULT
Neuronal Ash1l altered activity-dependent let-7b release and downstream cutaneous inflammatory signaling in the tested disease model.
SCIENTIFIC BOUNDARY
A psoriasis model does not establish psoriasis, neuroinflammation, or one shared skin mechanism in germline ASH1L haploinsufficiency.
Du et al., 2024 ↗10Direct tissue model
Skeletal muscle / myoblasts
Mouse myoblasts, developing muscle, and regeneration models
DIRECT RESULT
Ash1l activated Cdon and supported myoblast fusion, a process required for muscle formation, growth, and repair.
SCIENTIFIC BOUNDARY
The result does not establish a primary myopathy or explain every motor, tone, fatigue, or recovery finding.
Castiglioni et al., 2018 ↗11Direct tissue model
Bone / osteoclast lineage
Mouse osteoclast differentiation and bone-loss models
DIRECT RESULT
ASH1L restrained osteoclastogenesis and bone resorption in the tested models.
SCIENTIFIC BOUNDARY
Human bone fragility or osteoporosis is not established as an ASH1L-related-disorder mechanism.
Zhao et al., 2024 ↗12Dosage-context model
Fetal ovary / oocytes
Ash1l overexpression in mouse fetal ovaries
DIRECT RESULT
Excess Ash1l was associated with impaired DNA double-strand-break repair signaling and oocyte apoptosis in that model.
SCIENTIFIC BOUNDARY
The perturbation is overexpression—the opposite direction from haploinsufficiency—and cannot be used to infer fertility risk.
Zhang et al., 2022 ↗13Distal in vivo model
Heart development
Drosophila cardiac tissue
DIRECT RESULT
Ash1 with Caf1-55 and MRG15 supported H3K36me2-dependent heart development in vivo.
SCIENTIFIC BOUNDARY
A fly developmental result does not establish a human cardiac phenotype or clinical risk.
Zhu et al., 2023 ↗DISTINCT STATE QUESTIONS
Different experiments test different transitions and maintenance functions.
The evidence below concerns proliferation, differentiation, quiescence, inflammatory restraint, repair, resorption, or memory stabilization in different systems. “Failure to terminate” is not one mechanism; it must be decomposed into the specific entry, exit, maintenance, or recovery process that an experiment can measure.
Neural progenitor
Direct resultArrayed ASH1L knockout increased the proportion of Ki-67-positive progenitors.
BoundaryThis supports altered proliferative state; it does not directly prove a universal failure of neuronal differentiation.
Hematopoietic stem cell
Direct resultAsh1l-deficient cells failed to establish the normal quiescent adult stem-cell pool.
BoundaryFetal Sox17 and Lin28b programs were still extinguished, so the experiment does not show persistence of a global fetal state.
Epidermis
Direct resultAsh1l disruption altered the proliferation–differentiation balance, stratification, and wound re-epithelialization.
BoundaryA hypomorphic mouse-skin result is not proof of a shared human wound-healing mechanism.
Immune lineages
Direct resultSeparate models show roles in macrophage inflammatory restraint and induced regulatory-T-cell polarization.
BoundaryThey do not establish one human immune phenotype, occult inflammation, or immune-directed treatment.
Muscle and bone
Direct resultExperimental studies connect Ash1l to myoblast fusion and osteoclast differentiation or resorption.
BoundaryThese models do not establish that human weakness, hypermobility, fractures, or low bone density share one cause.
Memory circuit
Direct resultASH1L was required for later stabilization of selected memories in one thalamocortical mouse task.
BoundaryThis is maintenance in a defined circuit—not evidence of global human regression or neurodegeneration.
HUMAN CHRONOLOGY QUESTION
Do some states observed in this ASH1L corpus recover differently after physiological load?
The clinical corpus contains prolonged, partially reversible, and state-linked changes after illness, seizures, bowel or fuel burden, medication, anesthesia, pain, and hormonal transition. That pattern motivates prospective measurement of baseline, perturbation, duration, recovery curve, and tissue-appropriate readout. It does not establish one shared resolution defect or attribute every observed state to ASH1L.
CANCER-MODEL BOUNDARY
ASH1L is also studied as an oncogenic dependency. That is a different dosage problem.
ASH1L activity has been studied in KMT2A-rearranged leukemia and anaplastic thyroid-cancer models, often in settings of excess activity, overexpression, or tumor-specific dependency. Those experiments are valuable for catalytic and chromatin biology.
They do not establish that people with germline ASH1L haploinsufficiency have increased cancer risk, and an ASH1L inhibitor developed for leukemia is not a treatment rationale for ASH1L loss of function.
HUMAN RESEARCH AGENDA
Move from tissue plausibility to patient-variant mechanism.
The next step is not more organ-name association. It is paired human phenotyping and allele-aware measurement in the cell states capable of answering the question.
Brain, glia & memory
Measure allele-specific RNA/protein dosage in neurons and glia across differentiation; separate encoding, stabilization, retrieval, sleep state, and network excitability in human phenotyping.
Oral, enteric & GI
Build oral epithelial, enteric-neuron, smooth-muscle, and intestinal models only alongside objective feeding, motility, growth, or mucosal phenotypes.
Immune, skin & barrier
Pair defined clinical events with immune-cell, epithelial, and recovery assays. Systemic ESR/CRP and tissue-local function are different measurements.
Muscle & bone
Use standardized strength, fatigue, gait, fracture, mineral, and DXA phenotyping where indicated; then connect selected alleles to myoblast and osteoclast assays.
Endocrine & reproductive
Capture puberty, cycle, growth, and endocrine trajectories prospectively before selecting hormone-responsive cell models. Expression alone is not treatment evidence.
Cardiac, autonomic & renal
Start with reproducible human physiology—ECG, vitals, imaging, laboratory, or functional anchors—before assigning tissue-autonomous ASH1L mechanism.